FormalPara Key Summary Points

Why carry out this study?

Dry eye disease is known to affect corneal astigmatism measurement repeatability in preoperative patients with cataract.

Improvement in the consistency of corneal astigmatism measurements is essential for patients undergoing toric intraocular lens (IOL) implantation.

This research tested the hypothesis that dry eye treatment with long-acting diquafosol sodium (LA-DQS) has the potential to improve ocular surface condition and corneal astigmatism measurement repeatability, as well as to reduce refractive errors after cataract surgery.

What was learned from the study?

LA-DQS can improve preoperative corneal astigmatism measurement repeatability in patients with short tear break-up time (TBUT)-type dry eyes.

TBUT and high-order aberrations were significantly correlated with the reproducibility of preoperative corneal astigmatism measurements in short TBUT-type dry eyes.

These findings may enhance the development of methods improving IOL power calculation precision, particularly for toric IOLs, benefiting cataract surgery outcomes in patients with dry eyes.

Introduction

In recent years, there has been a notable increase in patient expectations regarding the accuracy of refractive results following cataract operations [1,2,3,4]. Irrespective of the type of intraocular lens (IOL) used, patients today seek not only visual clarity but also precise postoperative refractive outcomes [1]. Current scientific evidence shows that blurred vision accounts for 68% of the reports of patient dissatisfaction following cataract surgery, and the leading cause of blurred vision is postoperative refractive error (57%) [5].

The postoperative refraction of an IOL is calculated based on accurate measurements of multiple biometric variables, including anterior chamber depth (ACD), axial length (AL), and corneal curvature [6]. Of these factors, the most influential factor is corneal curvature since it accounts for approximately 70% of the refractive power of an optical system [6]. Therefore, the accuracy of corneal curvature measurement is essential to minimize postoperative refractive error. In addition, the use of toric IOLs, grounded in the accurate measurement of corneal astigmatism, is imperative for the effective improvement of both postoperative residual astigmatism and refractive error. In clinical practice, astigmatism originating from the lens can be corrected during cataract surgery, while that arising from the cornea can be mitigated via toric IOL implantation. A large research study evaluating the prevalence of corneal astigmatism in a sample of 15,448 patients reported a mean corneal astigmatism of 0.98 ± 0.78 diopters (D). Furthermore, 8% of the samples exhibited corneal astigmatism exceeding 2.00 D, while 2.6% had corneal astigmatism larger than 3.00 D [7]. Hence, it can be stated that a considerable number of patients with cataract stand to benefit from corneal astigmatism correction through toric IOL implantation. Therefore, precise preoperative corneal astigmatism measurements (power and axis), using well-established biometers such as the IOLMaster 700 (Carl Zeiss Meditec AG, Jena, Germany), are indispensable for the effective improvement of postoperative residual astigmatism.

However, prior studies have indicated that irregular astigmatism induced by dry eye can impact the precision and consistency of corneal astigmatism measurements [8,9,10,11,12], and approximately one-third of patients require further analysis to improve the precision of astigmatism measurements, even when well-established biometry devices are used [9]. Moreover, these studies have also demonstrated that dry eye induces instability in the tear film, leading to irregular astigmatism.

Trattler et al. indicated that about two-thirds of patients due for cataract surgery showed short tear breakup time (TBUT ≤ 5 s) [13]. In addition, Graae et al. reported that 55.5% of the Individuals designated to undergo cataract surgery experienced symptoms of dry eye [14]. As a result, many cataract surgery candidates are vulnerable to poor-quality preoperative corneal astigmatism measurements. Furthermore, in their review article, Venkateswaran et al. discussed the importance of preoperative dry eye treatment in patients scheduled for cataract surgery to obtain high-quality preoperative variable measurements [15]. However, there is a paucity of research exploring the effectiveness of dry eye treatment before cataract surgery on preoperative variable measurements. A previous study demonstrated that improvement of ocular surface condition improved repeatability of preoperative astigmatic measurements; however, this study used a different ophthalmic solution [10]. Although some studies have indicated that conventional 3% DQS improves postoperative ocular surface conditions in patients with cataract, no study has evaluated this effect in the preoperative period of cataract surgery [16, 17].

It has been reported that the efficacy of LA-DQS in improving ocular surface conditions in patients with dry eyes is comparable to conventional 3% DQS; however, this research was not conducted in the context of cataract surgery [18]. Therefore, as far as we know, the present study is the first to investigate the efficacy of LA-DQS on ocular surface condition and repeatability of astigmatic measurements in the preoperative period of cataract surgery. Furthermore, no previous study has investigated the effectiveness of dry eye treatment on the preoperative corneal astigmatism measurement repeatability using the power vector analysis. Thus, the outcomes of existing studies, highlighting the influence of dry eye disease on the repeatability of astigmatism measurements, coupled with the substantial prevalence of dry eye cases in preoperative patients with cataract, motivated us to proceed with the present study.

A 3% solution of diquafosol sodium (DQS) (Santen Pharmaceutical Co Ltd, Osaka, Japan) acts as a purinergic P2Y2 receptor agonist, stimulating the secretion of both mucin and tears [19, 20]. Further, it has been indicated that 3% DQS may enhance the dispersal as well as functionality of the tear film lipid layer [20]. In November 2022, a long-acting 3% diquafosol sodium (LA-DQS) ophthalmic solution (Santen Pharmaceutical Co. Ltd., Osaka, Japan) was released in Japan. LA-DQS contains polyvinylpyrrolidone (PVP) as an additive, resulting in an extended therapeutic effect of the ophthalmic solution. This can be attributed to two factors: the electrostatic interactions of PVP with secretory mucins, membrane mucins, and water, promoting the retention of tear fluid on the corneal surface [18], and the enhanced mucin secretion caused by DQS. As a result, the LA-DQS solution needs to be administered only three times a day in contrast to the six times daily application of DQS [18]. Therefore, anticipating a higher level of patient compliance, we decided to utilize LA-DQS in this study.

Our research was based on the hypothesis that dry eye treatment with LA-DQS has the potential to improve ocular surface condition and reproducibility of corneal astigmatism measurements, as well as to reduce refractive errors postoperatively. Thus, the aim of this research study was to assess the effect of LA-DQS on astigmatism measurement repeatability in preoperative cataract cases with short TBUT-type dry eyes, using power vector analysis. Further, we also examined the impact of LA-DQS treatment on the ocular surface condition of patients with dry eyes.

Methods

This multicenter, prospective, open-label study included 122 eyes of 61 patients and was performed at two private eye clinics, Yokosuka Chuoh Eye Clinic (31 patients) and Tsurumi Chuoh Eye Clinic (30 patients), between December 2022 and October 2023. Bilateral cataract surgery was performed in all patients diagnosed with dry eyes. This study was approved by the Yokosuka Chuo Eye Clinic and the Tsurumi Chuo Eye Clinic ethical committees (reference numbers: 2023-004 and 2023-005) and the data collection process adhered to the tenets of the Declaration of Helsinki of 1964. Written informed consent was obtained from all patients after providing a detailed explanation of the procedure and its possible outcomes. This research paper was written using the STROBE checklist for cross-sectional studies [21]. A post hoc power analysis indicated that 122 cases were required to detect a medium effect size of 0.5.

Participants

The ages of the studied participants ranged from 65 to 88 years. All participants were Japanese nationals. Of the 61 patients included in the final analysis, 30 were male and 31 were female. The present study utilized the Japanese dry eye diagnostic criteria, which incorporates TBUT ≤ 5 s and symptoms such as eye discomfort and visual disturbances to identify individuals with dry eyes [22]. Dry eye symptoms were diagnosed using the Japanese version of the Ocular Surface Disease Index (J-OSDI) [23]. However, the Japanese definition of dry eye does not include subjective symptom scores. Therefore, subjective symptoms were not quantified in this study.

Patients diagnosed with cataracts in both eyes (affecting their quality of vision and requiring surgery), those who met the Japanese dry eye diagnostic criteria in both eyes, strictly followed the LA-DQS treatment protocol and could maintain their head position in accordance with the examiner’s instructions were included in this study.

Patients who had received dry eye treatment and/or wore contact lenses within 1 month of study commencement were excluded from the analysis. Additionally, those who had used any eye drops within 24 h before the preoperative examinations, required any kind of eye drop therapy, or had active ocular or nasal allergies were also excluded. Furthermore, participants with a history of refractive procedures, ocular injury, ocular inflammation, ocular scarring, ocular dystrophy, or any other conditions that can cause irregularities of the ocular surface were not included in this study.

Measurement Device

All corneal astigmatism measurements were performed using the IOLMaster 700, which is a swept-source optical coherence tomography device with telecentric keratometry. The keratometer comprises 18 points arranged radially on three rings from the corneal center. Furthermore, the IOLMaster 700 includes a multi-dot telecentric keratometer, comprised of six dots in the central part of a 2.5 mm area. A camera detects the reflected spot patterns, and image separation is used to calculate the degree of corneal astigmatism [24].

Preoperative Dry Eye Treatment

LA-DQS was prescribed for administration exclusively to one eye (treatment group), while the fellow eye (non-treatment/control group) in the same patient remained untreated. In the treatment group, patients were instructed to use LA-DQS three times daily for 4 weeks before the preoperative examination. Patients were advised not to use any other eye drops before or during the study. The nurses confirmed patient compliance with the treatment protocol on the day of the preoperative examination. The eye drop administration protocol was designed based on a similar study that examined the impact of a different ophthalmic solution during the perioperative period of cataract surgery [25].

Examinations

Examinations of ocular surface conditions, TBUT, and corneal high-order aberrations (HOAs), were performed at baseline and 4 weeks after initiating dry eye treatment with LA-DQS in both the treatment and control groups. The TBUT was measured three times for up to 10 s each, and the mean value of the three TBUTs was calculated. The same experienced eye specialist checked TBUT at both clinics. The CASIA 2 device (Tomey Corporation, Nagoya, Japan) was used to check for HOAs within a 4-mm-diameter area centered on the cornea. The cylindrical power and meridian of astigmatism were measured on separate dates (3 and 4 weeks post-treatment) using the IOLMaster 700 in both groups. These measurements were performed once by the same qualified technician at each eye clinic. During the aforementioned examinations, the patients were asked to blink naturally while keeping their focus on the target device and to open their eyes normally to avoid their eyelids putting unwanted pressure on the eyeballs. Furthermore, participants were instructed to avoid unnecessary head movements during the examination. No eye drops were administered prior to the examination.

Statistical Analysis

All relevant variables were compared between the treatment and the non-treatment group. The Shapiro–Wilk test was used to check the normality of the variable distribution. Further statistical tests were performed based on the results of the normality tests. The baseline TBUT, HOAs, and ACD were analyzed using the Mann–Whitney U test. AL comparisons between the two groups were done using a paired t test. The Wilcoxon signed-rank test was used to analyze the changes in TBUT and HOAs between pre-(first check) and post-dry eye treatment (second check) in both groups.

The power vector method is the most reliable statistical tool for assessing both the shift in the meridian and the general trend simultaneously and quantitatively. This method can interpret the refractive errors of a spherocylindrical lens for graphical and numerical evaluations [26]. Furthermore, it is also useful for analyzing the statistical distribution of astigmatism [26]. Consequently, we employed the power vectors J0 and J45 to analyze the replicability of astigmatism measurements [26]. The values of these vectors are derived by employing a Fourier transformation which converts the corneal astigmatism from a sphero-cylinder notation to power vector notation:

$${\text{J}}0 = \left( {{\text{Cylinder}}/2} \right) \times \cos (2\;{\text{axis}}).$$
$${\text{J}}45 = \left( {{\text{Cylinder}}/2} \right) \times \sin (2\;{\text{axis}}).$$

In the above formulas, ‘cylinder’ denotes the positive cylindrical power, while ‘axis’ represents the minus of cylinder's meridian. ‘J0’ signifies the cylinder power at the orthogonally located 90° and 180° meridians. A positive ‘J0’ value indicates with-the-rule astigmatism, whereas a negative value shows against-the-rule astigmatism. ‘J45’ values indicate a cross-cylinder set positioned at the 45°and 135° meridians, indicating oblique astigmatism [26].

The effective astigmatic cylindrical power was evaluated based on the results for J0 and J45 [26]. The repeatability of the astigmatism measurements was described as the within-subject standard deviation (Sw). Repeatability was evaluated as the square root of the mean square within groups, using a one-way ANOVA test [27]. F-test was used to compare the repeatability of J0 and J45 (J0 Sw and J45 Sw).

In the control group, the correlation of J0 Sw and J45 Sw with TBUT and HOAs at baseline was analyzed using Spearman’s rank-correlation coefficient. Additionally, the relative effects of TBUT and HOAs on J0 Sw and J45 Sw were analyzed using a multiple linear regression model (stepwise variable selection). Furthermore, the association between the effective astigmatic cylindrical power evaluated from the results of J0 and J45, and its repeatability, represented as the Sw, was analyzed using Spearman’s rank-correlation coefficient.

All statistical analyses were performed using the SPSS software (v. 25.0; IBM SPSS Statistics, Armonk, NY, USA). P < 0.05 was considered to be statistically significant. A post hoc analysis was performed using G* Power (v. 3.1.9.7) to check the sample size power.

Results

A total of 122 eyes of 61 patients satisfied the inclusion criteria and were analyzed in the present study. All participants used LA-DQS for 4 weeks before the preoperative examination. No adverse drug reactions were observed during this research. The given sample size (n = 122) demonstrated a robust statistical power in detecting medium effect sizes (d = 0.5), as indicated by the results of the post hoc analysis. The power of the F-test was 97.6% for J0 and 86.9% for J45, respectively.

Table 1 shows the baseline data of the eyes in the treatment and control groups, indicating no significant differences between the two groups, at the outset of the study.

Table 1 Baseline characteristics of the two study groups
Full size table

Table 2 compares the averages of J0 and J45 between the two study groups, which were measured on separate dates. Both J0 and J45 averages showed no significant differences between groups (P = 0.514, P = 0.738).

Table 2 Comparison of average J0 and J45 values between the two study groups
Full size table

Table 3 shows the differences in TBUT and HOAs between pre-(first check) and post-dry eye treatment (second check) for both groups. Both TBUT and HOAs showed significant changes in the treatment group (P < 0.001 and P < 0.001, respectively); however, these changes were not significant in the non-treatment group (P > 0.999 and P = 0.908, respectively).

Table 3 Changes in TBUT and HOAs between pre-(first check) and post-dry eye treatment (second check) in both study groups
Full size table

Table 4 shows the reproducibility of J0 and J45 in the treatment and control groups, indicated as the within-subject standard deviations (J0 Sw and J45 Sw). A significant difference was observed in J0 Sw and J45 Sw between the treatment and control groups (P < 0.001 and P = 0.002, respectively).

Table 4 Repeatability of J0 and J45 within the two study groups
Full size table

Figure 1 demonstrates the correlations of J0 Sw and J45 Sw with TBUT and HOAs at baseline in the control group (Fig. 1a–d). J0 Sw and J45 Sw showed significant correlations with TBUT (rho [ρ] = − 0.481, P > 0.001; ρ = − 0.751, P > 0.001) and HOAs (ρ = 0.568, P > 0.001; ρ = 0.730, P > 0.001). When P < 0.0001, adjusted R2 = 0.333, variance inflation factor (VIF) = 1.000, and Durbin–Watson ratio = 1.994, HOAs showed a stronger relative association with J0 Sw than with TBUT. When P < 0.0001, adjusted R2 = 0.571, VIF = 2.944, and Durbin–Watson ratio = 1.784, HOAs showed a stronger relative association with J45 Sw than with TBUT.

Fig. 1
figure 1

a–d Correlations of J0 Sw and J45 Sw with TBUT and HOAs at baseline in the non-treatment group. a Correlation between J0 Sw and TBUT. b Correlation between J0 Sw and HOAs. c Correlation between J45 Sw and TBUT. d Correlation between J45 Sw and HOAs. TBUT tear break-up time, HOAs high-order aberrations, Sw within-subject standard deviation

Full size image

Figure 2 shows the correlation between the effective astigmatic cylindrical power evaluated from the results of J0 and J45 (repeatability indicated by the Sw). No significant correlation was observed (ρ = 0.166, P = 0.197).

Fig. 2
figure 2

Correlation between the effective astigmatic cylindrical power calculated from the results of J0 and J45 (magnitude of astigmatism) and its repeatability expressed as within-subject standard deviation (magnitude of astigmatism Sw). Sw within-subject standard deviation

Full size image

Discussion

The primary finding of this study was that preoperative treatment with LA-DQS improved the repeatability of astigmatism measurements in cataract surgery cases involving dry eyes. Additionally, as far as we know, this is the first study investigating the effect of dry eye treatment on the preoperative corneal astigmatism measurement repeatability using appropriate statistical analysis methods.

Previous studies have explored the effectiveness of conventional 3% DQS in improving the tear film stability in patients with dry eyes. Kaido et al. demonstrated that TBUT and HOAs in short TBUT-type dry eyes significantly improved with conventional 3% DQS therapy [28]. Koh S. also indicated a significant effect of conventional 3% DQS on the improvement of tear film stability in various types of dry eyes, including short TBUT-type dry eyes [20]. However, it has been shown that only 8.3% patients strictly follow the recommended administration schedule (six times a day) [29], indicating that the patients fail to derive full benefit of the treatment due to poor compliance.

Unlike 3% DQS, LA-DQS eye drops need only three applications per day. In addition, two previous studies have reported that the performance of LA-DQS is comparable to that of 3% DQS [29, 30]. Therefore, we decided to use LA-DQS in the present study, with the expectation of a higher level of patient compliance [18, 30].

A multitude of previous studies have investigated spherocylindrical refractive errors using mathematical representations and statistical manipulations. There has been much debate about the difficulties encountered when astigmatism is shown in the conventional polar form of magnitude and axis instead of the more mathematically tractable Cartesian form [31]. Generally, the statistical analysis of angular data such as astigmatism axes is essentially different from the analysis of nondirectional data [31]. Therefore, misleading results can be a consequence of the inappropriate application of traditional statistical methods to nondirectional data analysis. Therefore, this study used a power vector analysis for astigmatism calculations.

Our findings revealed significant positive effects of LA-DQS on ocular surface conditions, TBUT, and HOAs in preoperative patients with cataract and dry eyes. These results align with a previous study that evaluated the impact of conventional 3% DQS on HOAs in patients with dry eyes [16].

The novel finding of the present study was the improvement in the replicability of astigmatism measurements in preoperative patients with cataract and dry eyes after treatment with LA-DQS. Previous studies have indicated the negative impact of dry eyes on the repeatability of keratometry measurements [12, 32]. Epitropoulos et al. reported hyperosmolarity of the tear film as one of the factors in the definition of dry eye in the International Dry Eye Workshop (DEWS) and demonstrated the negative effects of tear hyperosmolarity on the repeatability of keratometry measurements in patients scheduled for cataract surgery [32]. Furthermore, Hiraoka et al. demonstrated the significant impact of short TBUT on corneal astigmatism measurement repeatability [12]. Similarly, the present study showed that short TBUT-type dry eye aggravated the repeatability of corneal astigmatism measurements, which was improved by dry treatment with LA-DQS due to tear film stabilization. TBUT is the principal index of tear film instability, which can increase HOAs [33]. Based on this theory, the ocular surface condition in individuals with dry eyes can negatively affect the accuracy of astigmatism measurements [34]. These derivations are further supported by the findings of the present study, which reveal that TBUT and HOAs have a substantial correlation with the reproducibility of astigmatic measurements (J0 Sw and J45 Sw).

Since the accuracy of corneal astigmatism measurements may be reduced in patients with dry eyes or tear film instability and LA-DQS can improve the tear film integrity, we can speculate that the replicability of astigmatism measurements in the present study showed significant improvement due to the stabilization of the tear film by LA-DQS.

The present study also showed that TBUT and HOAs correlated significantly with the consistency of astigmatism measurements. Previous research has also demonstrated a significant correlation between TBUT and the repeatability of corneal curvature measurements [12]. In addition, our data revealed that HOAs had a stronger relative association with the replicability of astigmatism measurements than TBUT. Some research papers have reported an association between TBUT and HOAs in dry eyes [35, 36], and both the parameters have been shown to improve by stabilization of the tear film with secretagogues (eye drops), such as 3% DQS and rebamipide [33, 37]. Although these results support our findings, we were unable to comment on the detailed mechanism behind the stronger relative association of HOAs with TBUT, as various factors beyond our current scope of investigation may influence HOAs in dry eyes [38].

Another novel finding of the present study was that even minimal levels of astigmatism can lead to a lack of consistency in preoperative astigmatism measurements.

Managing astigmatism has become increasingly important with the growing popularity of multifocal IOLs [39,40,41]. Astigmatism levels higher than 1.00 D require correction in order to gain better vision with diffractive multifocal IOLs [40]. Currently, the superior visual quality offered by diffractive trifocal IOLs compared to bifocal IOLs has led to an increase in their demand [42, 43]. However, the quality of vision with trifocal IOLs is more vulnerable to residual astigmatism [39]. Since astigmatism above 0.75 D requires correction to obtain improved visual outcomes [39], minimizing postoperative residual astigmatism is crucial to increase patient satisfaction with diffractive trifocal IOLs. This means that it is crucial to improve the quality of preoperative astigmatism measurements regardless of the magnitude of astigmatism and to carefully select the appropriate power and axis when using toric IOLs.

Additionally, the American Society of Cataract and Refractive Surgery (ASCRS) Cornea Clinical Committee created a new consensus-based practical diagnostic ocular surface disease (OSD) algorithm to address educational gaps between their members and the committee [44]. In this algorithm, the authors insisted on the significant influence of tear film instability caused by OSD on the accuracy of preoperative keratometry measurements. They emphasized that surgeons should take this into account, particularly in patients with presbyopia-correcting IOLs (multifocal or accommodating IOLs). The results of the aforementioned studies, as well as our own, support this idea.

This study has some limitations. First, the Japanese dry eye definition was applied to identify patients with dry eyes, which differs in certain aspects from the other scientific classifications used around the world. In addition, the Japanese definition of dry eye does not include subjective symptom scores. Therefore, subjective symptoms were not quantified in this study. Thus, it is important to examine the same hypothesis using alternative definitions in future studies, given the potential for different results based on varying criteria. Second, we did not investigate the clinical effects of improvement in the repeatability of astigmatism measurements on the selection of toric IOLs. Due to the possible involvement of other factors in IOL calculations, improved astigmatism measurement repeatability may not be sufficient to improve IOL calculation accuracy. Third, we did not investigate the scientific mechanism underlying the correlation between TBUT, HOAs, and astigmatism measurement repeatability. This should be investigated further in future studies.

Conclusions

This study revealed that dry eye treatment with LA-DQS significantly improved ocular surface condition and corneal astigmatism measurement repeatability in preoperative patients with cataract and short TBUT-type dry eyes. Our results also indicate the importance of preoperative dry eye treatment regardless of the magnitude of astigmatism. The findings of this study may contribute to the development of new methods and techniques that can improve IOL power calculation precision in patients with dry eyes, especially when considering the application of toric IOLs. Additionally, our findings may help improve the accuracy of postoperative refraction of cataract surgery in patients with dry eyes. In the future, improvement of postoperative residual astigmatism in toric IOL implantation cases with dry eyes treated preoperatively with LA-DQS should be investigated to analyze the relationship between the accurate selection of the toric IOL style and astigmatism measurement repeatability.